A combined fuzzy and least squares method approach for the evaluation of management questionnaires

L. Kóczy, Ojaras Purvinis, Dalia Susnienė

Research output: Chapter in Book/Report/Conference proceedingChapter


A set of answers to questions to employees of various companies in Lithuania may refer to various positive and negative aspects of the attitudes of employees. These are called Organizational Citizenship Behavior (positive) and Counterproductive Work Behavior (negative). The components in the answers may be grouped by expert knowledge, and by statistical analysis and, according to these approaches, based on expert domain knowledge by management specialists, fuzzy signature structures describing the mutual effects of single elements in the questionnaire may be created. There are some slight differences between the two results, that indicate that expert knowledge is sometimes not objective. An additional step applying hybrid Generalised Reduced Gradient algorithm and Genetic Evolutionary Algorithm for heuristic optimization of the aggregation parameters in the Fuzzy Signatures reveals a final model according to the responses. These latter results raise some new questions, including the idea of the use of undeterministic graphs, thus resulting in Fuzzy Fuzzy Signatures. The method could be applied to other similar multicomponent vague data pools.

Original languageEnglish
Title of host publicationStudies in Computational Intelligence
PublisherSpringer Verlag
Number of pages9
Publication statusPublished - Jan 1 2020

Publication series

NameStudies in Computational Intelligence
ISSN (Print)1860-949X



  • Fuzzy signature
  • Least squares method
  • OCB

ASJC Scopus subject areas

  • Artificial Intelligence

Cite this

Kóczy, L., Purvinis, O., & Susnienė, D. (2020). A combined fuzzy and least squares method approach for the evaluation of management questionnaires. In Studies in Computational Intelligence (pp. 157-165). (Studies in Computational Intelligence; Vol. 819). Springer Verlag. https://doi.org/10.1007/978-3-030-16024-1_20